There were 220 patients studied, of which 44 were considered "critical" according to pre-established factors. In all of them, intra-gastrical pH in serious forms was detected, establishing that 45% of the "non-critical" patients had a pH factor less than 4. In these 61 patients and in the 44 "critical" patients a gastric alkalinization was sought, using between 1.2 and 2.4 grams of cimetidina daily, to which alkalines were added when necessary. Nine percent of the critical patients (4 of 44) had H.D.A., while 11 of 33 (33%) of the critically ill from a previous series that had not been treated had pronounced gastro-intestinal bleeding lesions (p less than 0.01.) The potential difference of gastric mucous in critical patients was measured likewise, being significantly less (means = 17.37 +/- 1.13) than in the non-critical patients (means = 30.2 +/- 1.13) (p less than 0.005). The use of cimetidina as an inhibitor of gastric secretion in a dosage of 1.2 grams was effective in 73% of the non-critical patients and in only 29% of the critical patients. Despite an increase in the dosage up to 2.4 grams in 9% of the critical and 22% of the non-critical patients, it was necessary to add alkalines. This lack of response shows a correlation between the amount of risk factors (3 of 10 patients had 3 factors) and the existence of hepatitic insufficiency (5 out of 16 patients and 5 out of 26 of sepsis).